used 12at7

[pforeman]

I've been looking at used 12at7 tubes and see them advertised as having a reading of (for example ) 4725 / 4400 (on a Hickok tester).
Is there a value that relates to 100% ( NOS ) ?
Is there a value that relates to "don't buy it " - it's only got a few months left
Is there a value that relates to "used" but lots of life left?

In the tube data sheets I see several values for Ra from 10,500 to 16,500 depending on Va and Vg1.

Thanks
Paul

 

[SY]

Why not just get a new one? The JJs are the best 12AT7/ECC81 ever made and they're something like $10.

 

[pforeman]

Hi Sy,
Yes, I have a J & J and and old Amperex Bugle Boy.
No way to test the Bugle Boy to find out how "old" it is. Just thought that it had a little more believable sound stage.
I was originally trying to find the best output tubes (Tubelab S SE -triode only) but someone posted that it was the input tube that mattered most, so I finnegled a old Bugle Boy from a friends guitar amp...
I have a J&J, so just wanted to see for not too much money, I can buy a used 12at
or a new siemens (about the same $)
For $25 or $30 I thought I could try one of the "venerated" tubes, I just don't understand the results when posted about micro-ohms (?).
Thanks
Paul

 

[DF96]

It is more or less impossible to tell by measurements how much use a valve has had. The raw gm figure (presumably what is being quoted) varies quite a bit between guaranteed unused ones.

Measuring the change in current or gm when the heater voltage is reduced may give an indication that a valve is nearing the end of its emission life. A high ionic grid current shows a gassy valve.

Two options: buy new ones from a reliable dealer, or buy cheap NOS ones off the internet and expect to throw away some of them. Let other people buy the expensive NOS ones.

 

[SY]

Maybe micromhos? It's a measure of transconductance, but not meaningful without knowing current and voltages. I've tested most of the "venerated" ECC81 types and the JJ was by a large margin the most linear and microphonic-free.

 

[DF96]

One of the mysteries of the universe is why Americans measure gm in micro-mhos, but Europeans measure it in mA/V (=milli-mhos). 4400 micro-mhos = 4.4 mA/V.

Was the linear JJ 12AT7 the frame-grid version? I would expect a frame-grid to give greater linearity, if done properly, as it means the valve is basically planar as far as the electrons are concerned and this should lead to constant mu.

 

[SY]

We have a finer discernment than Brits.

 

[Salas]

Or milli Siemens. mS.

 

[DF96]

I think you are just afraid of decimal points, so prefer integers.

 

[Salas]

SI is Siemens. Backwards spelled Ohm in micro or milli is not for datasheets anymore. Japanese Jfet datasheets state in mS for instance. Which is mA/V. Even the Russian tube datasheets have mA/B.

 

[SY]

Apparently Russians can't spell "bolt."

 

[el156]

One of the mysteries of the universe is why Americans measure gm in micro-mhos, but Europeans measure it in mA/V (=milli-mhos). 4400 micro-mhos = 4.4 mA/V.

Was the linear JJ 12AT7 the frame-grid version? I would expect a frame-grid to give greater linearity, if done properly, as it means the valve is basically planar as far as the electrons are concerned and this should lead to constant mu.

There is no mistery in the American measure of GM, 4400 micro-mhos is far more precise than 4,4 ma/v

 

[Salas]

Apparently Russians can't spell "bolt."

''B'' is pronounced as ''V'', it is the Greek 'beta' letter. Greeks we pronounce it ''vita''. Russians got alphabet from two Byzantium era Greek monks from Salonica. One's name was Cyril. Hence they use Cyrillic. The Byzantine two headed eagle facing east and west is their emblem still, Moscow was symbolically the third Rome after the fall of Constantinople and Czar meant Caesar.

 

[DF96]

4400 and 4.4 both contain two significant figures, so possibly same precision, except 4400 is ambiguous as it could mean four sig figs so it actually conveys less information. GM make automobiles, gm measures valves.

 

[SY]

4725 is four significant figures. But then again, our electricals weren't designed by Lucas.

 

[DF96]

If I wanted to be awkward (moi?) I could say that 4725 might be about 2.5 sig figs. 4729 would definitely be 4 sig figs. Anyway, this is getting well OT so one of the mods will be slapping our wrists soon. Wait a minute: SY is a mod, so that's all right then. Have we answered the OPs question?

 

[trobbins]

Is there a value that relates to 100% ( NOS ) ?
- Not when just looking at one ebay item which gives one figure.

Is there a value that relates to "don't buy it " - it's only got a few months left
- Not when just looking at one ebay item which gives one figure. But there is confidence that the valve works as it is expected.


Is there a value that relates to "used" but lots of life left?
- Not when just looking at one ebay item which gives one figure.

Valve testers in general do not provide comparable measurements with a high level of accuracy (as you may expect of other meters of different manufacturers). They weren't subject to an international certification scheme for accuracy, and they don't come with a golden sample that you can pass around and use as a comparison standard. As such, they tell you the valve is working based on valve tester expected range of values for a tube, and if you test many on the same tester over time then confidence builds that a particular valve is behaving in a normal manner. It gets complex when using them for more detailed assessments.

Ciao, Tim